walrus01

44 ms·

661.
by walrus01 3mo ago
Toolcraft: https://www.toolcraft.com/ One of the better medium volume CNC shops around.
662.
by walrus01 3mo ago
It's my understanding that small to medium sized cities (and rural areas) depopulating is going on now throughout almost all of Japan. One of the things that really surprised me recently is how low the cost of real estate is for an old
663.
by walrus01 3mo ago
It reminds me a little bit of a Phrack textfile, but if Phrack had raster image art advertising in it.
664.
by walrus01 3mo ago
Cue the Steve Ballmer "developers developers developers developers" video. Bonus points if extra sweaty.
665.
by walrus01 3mo ago
a relatively detailed analysis of estimated radiation for a 650 day trip to mars: https://pmc.ncbi.nlm.nih.gov/articles/PMC9916691/ from other sources If I'm reading things right, one estimate of a round trip
666.
by walrus01 3mo ago
Since Pete took the helm and renamed things, it's definitely now the Whiskey Advanced Research Projects Agency.
667.
by walrus01 3mo ago
It sounds like each "MEP" (mission extension pod) once attached is its own fully autonomous satellite? How big is each MEP? Like, a mini fridge? Since the mission is designed to attach one and then leave to go to another satellite
668.
by walrus01 3mo ago
Presumably it's also known to everyone who builds/operates these that having a foreign nation's military satellite intentionally approach your own geostationary satellite to within a very short distance, as the vast area up t
669.
by walrus01 3mo ago
The new robot grappler arm definitely seems like an advancement on previous designs. The MEV was, as I recall, designed to mate with only a specific set of models of geostationary satellite bus and, for lack of a better term, ram a probe up
670.
by walrus01 3mo ago
One of the problems in buying used/refurb x86-64 rack servers for test and development/proof of concept environment, is that by volume in the market, there's not that many -48VDC power supplies going around, because maybe 5-1
671.
by walrus01 3mo ago
> Are there? At the highest levels of defense and law, AWS and Azure are used. This is certainly true if the user is an American company. You could look at the European initiatives to run this stuff on hardware they own in facilities the
672.
by walrus01 3mo ago
You seem to be categorizing everything that considers their data being in the possession of the US an unacceptable risk to be tinfoil hat, which is kind of an insult to a large portion of the world. If you haven't been paying attention
673.
by walrus01 3mo ago
It's not exactly the first, this is the first (unclassified, that we know about) satellite to go up and grapple a satellite in geostationary and provide stationkeeping functions for it, which did so back in 2020: https://en.
674.
by walrus01 3mo ago
Looks like it's live now, it's approx 17GB per safetensors file x 96 files, so so about 1.63TB. I can only imagine that people with their favorite quantizing tools warmed up and ready to go are aggressively downloading it now.
675.
by walrus01 3mo ago
I am not in Quebec but Quebec hydro rate D for standard residential would be one example of around what I pay. https://www.hydroquebec.com/residential/customer-space/rates... Another example would be Manitoba hydr
676.
by walrus01 3mo ago
The epyc Venice SP7 socket is apparently 9324 pins https://x.com/tomshardware/status/2066846693778510331
677.
by walrus01 3mo ago
It's a bit worrying the author is asking chatgpt about the tingle rather than making sure there's a grounding lug on all the panels and they're all bonded to a grounding rod. The rudimentary nature of the setup makes me think
678.
by walrus01 3mo ago
Very good description of it. It does seem like a bit of a rhetorical question to ask a forum that has a very high population of Linux and BSD users why they might desire to have the option to do something themselves rather than relying on a
679.
by walrus01 3mo ago
It is not my use case but given recent political developments in international relations caused by the executive branch of the US government, off the top of my head, I could think of a lot of European or Canadian firms for which that would
680.
by walrus01 3mo ago
1536GB of DDR4 ECC server RAM is somewhere between $4000-6000 USD used right now, by the time you put in parallel enough NVME SSD to approach good speeds, you'd be approaching that (and also likely running out of PCI-E bus lanes direct
681.
by walrus01 3mo ago
Keep in mind that just because it has dual 750W power supplies that doesn't mean it's what its load will be, for a full CPU loaded wattage figure you'd need basically a pair of kill-a-watts plugged in inline on the feed for e
682.
by walrus01 3mo ago
I think he means electrical bill at his estimated wattage load of the server and his known kWh cost, not rented server/hosting cost.
683.
by walrus01 3mo ago
I have never seen anyone report "this produced really great results" from intentionally quantizing their context vs. leaving it at full precision which is the ordinary default.
684.
by walrus01 3mo ago
Yeah, or close enough to 5GB for estimation purposes, for example a just spawned qwen 3.5 122B llama-server instance reports as: 0.07.015.888 I common_memory_breakdown_print: | - Host | 170038 = 162913 + 6
685.
by walrus01 3mo ago
The cynic in me says maybe they would be further along if they hadn't spent $80 billion on trying to build the "Metaverse" VR world. I've never met anyone who actually uses it and to the best of my knowledge it has very
686.
by walrus01 3mo ago
Do you mean by trading dollars for the privacy you need as: a) Contracting with a third-party independent inference provider who will run your choice of model on fast hardware that they own, with all appropriate data security/privacy&#
687.
by walrus01 3mo ago
Recently spent a few hours messing with bonsai 27B and ternary bonsai 27B at total weights + context fitting in slightly under 6GB RAM, and it's just dumb as hell. It writes what seems like grammatically correct content but it's e
688.
by walrus01 3mo ago
Or 11/Shrawan 2083 if you're in Nepal
689.
by walrus01 3mo ago
I assume you mean putting only the 32B active parameters on the GPU, and the rest on a bunch of regular server DRAM like on a 768GB to 1024GB RAM server? Because Kimi K2.6 in Q4 is about 584GB GGUF size on disk and will use slightly more
690.
by walrus01 3mo ago
It's a great concept but I think it would cross the line from 'very slow' to 'so slow it's unusable' at this size. Even if we say you have an NVME SSD that does 7GB/s reads, that's dramatically slower

More ›